Variations in composition of the carbohydrate, saccharide, or sugar molecule, have been shown to affect the affinity of IgG for three classes of FcγRs (FcγRI, FcγRII, and FcγRIII) that link IgG-mediated immune response with cellular effector functions (Wright and Morrison, Trends Biotechnol 15(1): 26-32; Gessner et al., Ann Hematol 76(6):231-48 (1998); Jefferis et al., Immunol Rev 163: 59-76 (1998). Ravetch and Bolland, Annu Rev Immunol 19: 275-90 (2001)).
Sugar chains of glycoproteins are generally divided into the following two broad types based on the binding form to a proteinaceous moiety: namely a sugar chain which binds to asparagine (N-glycoside-linked sugar chain), and a sugar chain which binds to other amino acids, such as serine or threonine (O-glycoside-linked sugar chain). Typically, they have a basic common core structure shown by the following structural formula (I):

The N-glycoside-linked sugar chains have various structures, with various sugar molecules. The sugar chain terminus which binds to asparagine is typically called a reducing end, and the opposite side is called a non-reducing end. N-glycoside-linked sugar chains can include a high mannose type in which mannose alone binds to the non-reducing end of the core structure; a complex type in which the non-reducing end side of the core structure has at least one parallel branch of galactose-N-acetylglucosamine (Gal-GlcNAc) and the non-reducing end side of Gal-GlcNAc has a structure of sialic acid, bisecting N-acetylglucosamine or the like; a hybrid type in which the non-reducing end side of the core structure has branches of both of the high mannose type and complex type; and the like. The structure of a sugar chain can be determined by sugar chain genes, such as a gene for a glycosyltransferase which synthesizes a sugar chain, and/or a gene for a glycolytic enzyme which hydrolyzes the sugar chain.
Glycoproteins are typically modified with a sugar chain in the endoplasmic reticulum (ER) lumen. For example, during the biosynthesis step of the N-glycoside-linked sugar chain, a relatively large sugar chain is transferred to a polypeptide chain that is elongating in the ER lumen. Sugar molecules can be added in succession to phosphate groups of a long chain lipid carrier comprising about 20 α-isoprene units, such as dolichol phosphate (P-Dol). For example, N-acetylglucosamine (GlcNAc) is transferred to P-Dol to form GlcNAc-P-P-Dol and then one more GlcNAc is transferred to form GlcNAc-GlcNAc-P-P-Dol. Next, five mannoses (Man) are transferred to thereby form (Man)5-(GlcNAc)2-P-P-Dol and then four Man's and three glucoses (Glc) are transferred. As a result, a sugar chain precursor, (Glc)3-(Man)9-(GlcNAc)2-P-P-Dol, a core oligosaccharide, is formed. The sugar chain precursor comprising 14 sugars can then be transferred to a polypeptide having an asparagine-X-serine or asparagine-X-threonine sequence in the ER lumen. The dolichol pyrophosphate (P-P-Dol) bound to the core oligosaccharide is typically released and becomes dolichol phosphate (P-Dol) by hydrolysis with pyrophosphatase, and is recycled. Trimming of the sugar chain typically starts after the sugar chain binds to the polypeptide. For example, 3 Glc's and 1 or 2 Man's are eliminated on the ER, such as by the action of α-1,2-glucosidase I, α-1,3-glucosidase II and α-1,2-mannosidase.
The glycoprotein which was subjected to trimming on the ER can be transferred to the Golgi body and further modified. For example, present in the cis part of the Golgi body are N-acetylglucosamine phosphotransferase (which aids in the addition of mannose phosphate), N-acetylglucosamine 1-phosphodiester α-N-acetylglucosamimidase and α-mannosidase I (which reduce the Man residues to 5). Present in the medium part of the Golgi body are N-acetylglucosamine transferase I (GnTI) (which aids in the addition of the first outside GlcNAc of the complex type N-glycoside-linked sugar chain), α-mannosidase II (which aids in the removal of 2 Man's), N-acetylglucosamine transferase II (GnTII) (which aids in the addition of the second GlcNAc from the outside) and α-1,6-fucosyltransferase (which aids in the addition of fucose to the reducing end N-acetylglucosamine). Present in the trans part of the Golgi body are galactose transferase, which aids in the addition of galactose, and sialyltransferase, which relates to addition of sialic acid such as N-acetylneuraminic acid or the like. Thus, various N-glycoside-linked sugar chains can be formed by activities of these various enzymes.
Sugar chain structure variations, or variant glycosylation patterns due to various sugar molecule content in such chains, plays an important role in the effector function of glycoproteins, such as antibodies. For example, in the Fc region of an antibody of an IgG type, two N-glycoside-linked sugar chain binding sites are typically present. In serum IgG, the sugar chain binding site generally binds a complex type sugar chain having multiple branches, and in which addition of sialic acid or bisecting N-acetylglucosamine is low. There are a variety of manners and forms in which addition of galactose is made to the non-reducing end of the complex type sugar chain and the addition of fucose to the N-acetylglucosamine in the reducing end (see for example, Leppanen et al., Biochemistry, 36, 7026-7036 (1997)).
Fucosylation is an example of a process in which newly synthesized antibodies can be modified by the addition of fucose saccharides in the Golgi apparatus of a cell. This protein modification can be visualized by staining the cells with fluorophore-conjugated LCA (Lens culimaris agglutinin-A), a chemical that preferentially binds to proteins modified with fucose. Recently, it has been observed that fucosylation of antibodies affects antibody binding to human FcγR and antibody-dependent cellular cytotoxicity (ADCC). Antibody-binding affinity and antibody-mediated ADCC is strongly enhanced when antibodies have low levels of fucose (Shields et al., J Biol Chem 277(30): 26733-4 (2002); Shinkawa et al., J Biol Chem 278(5): 3466-73 (2003)).
Protein fucosylation is a process that begins with the uptake of free fucose, followed by phosphorylation by fucose kinase and conversion to GDP-fucose by GDP-fucose pyrophosphorylase. Fucosyltransferases transfer the fucose residue to glycans or protein within secretary pathways, subsequently the modified glycoproteins are delivered to the cell surface for secretion. The fucosylation status of antibody-producing cells correlates with the fucose content in the antibody produced, and that the absence of fucosyltransferase abrogated the fucosylation at both cellular and antibody levels (Yamane-Ohnuki et al., Biotechnol Bioeng 87(5): 614-22 (2004)).
ADCC is an important mechanism of action by which therapeutic antibodies induce immune responses and mediate the killing of cancer cells. Enhancement of ADCC by therapeutic antibodies can improve clinical responses and reduce the therapeutic dosages, thus diminishing possible side effects (Adams and Weiner, Nat Biotechnol 23(9): 1147-57 (2005)). In vivo models and clinical trials have demonstrated that therapeutic antibodies, such as Herceptin, possess cytotoxic properties, including ADCC. These properties are main factors in Herceptin induced breast tumor regression and protection from lung metastasis (Carter et al., Proc Natl Acad Sci USA 89(10): 4285-9 (1992); Lewis et al., Cancer Immunol Immunother 37(4): 255-63 (1993); Cooley et al., Exp Hematol 27(10): 1533-41 (1999); Clynes et al., Nat Med 6(4): 443-6 (2000); Repka et al., Clin Cancer Res 9(7): 2440-6 (2003); Gennari et al., Clin Cancer Res 10(17): 5650-5 (2004); Nahta and Esteva, Cancer Lett 232(2): 123-38 (2006)). Further, it was demonstrated that antibodies produced by fucosylation-low cells enhance ADCC activity (Shields et al., J Biol Chem 277(30): 26733-4 (2002); Shinkawa et al., J Biol Chem 278(5): 3466-73 (2003)).
Expression of ADCC activity of human IgG1 subclass antibodies typically requires binding of the Fc region of an antibody to an antibody receptor existing on the surface of an effector cell, such as a killer cell, a natural killer cell, an activated macrophage or the like (FcγR) and various complement components. It has been suggested that several amino acid residues in the second domain of the antibody hinge region and C region (hereinafter referred to as “Cγ2 domain”) and a sugar chain linked to the Cγ2 domain are important for this binding reaction.
Currently, several strategies have been proposed for enhancing monoclonal antibody-mediated ADCC against tumor cells, such as: 1) developing specific anti-cancer antibodies in which one arm of the antibody binds to an IgG receptor in order to more efficiently recruit immune effector cells (Segal et al., J Immunol Methods 248(1-2): 1-6 (2001)); 2) using recombinant human cytokines to increase the effector function of immune effector cells (Carson et al, Eur J Immunol 31(10): 3016-25 (2001); Repka et al., Clin Cancer Res 9(7): 2440-6 (2003)); 3) using IgG-cytokine fusion protein (Penichet and Morrison, J Immunol Methods 248(1-2):91-101 (2001)); 4) altering the Fc sequence of an antibody for improved binding to an IgG receptor (Shields et al., J Biol Chem 276(9): 6591-604 (2001)); and 5) optimization of the levels of Asn297-linked carbohydrates (Umana et al., Nat Biotechnol 17(2): 176-80 (1999); Davies et al., Biotechnol Bioeng 74(4): 288-94 (2001); Shinkawa et al., J Biol Chem 278(5): 3466-73 (2003)).
One approach is to modify the fucose content of anti-cancer antibodies to increase binding affinity for FcγRs and ADCC. IgG1 has two N-linked oligossacharide chains bound to Asn297, composed of a trimannosyl core structure with the presence or absence of a core fucose, bisecting N-acetylglucosamine and terminal galactose (Rademacher et al., Biochem Soc Symp 51: 131-48 (1986)). The nature and importance of Asn297-linked carbohydrates in immunoglobulin G effector functions has long been recognized. It has been demonstrated that defucosylated Rituxan, an anti-CD20 antibody for lymphoma treatment, strongly binds to FcγRIIIa with high affinity and 100-fold enhanced ADCC activity (Shinkawa et al., J Biol Chem 278(5):3466-73 (2003); Yamane-Ohnuki et al., Biotechnol Bioeng 87(5): 614-22 (2004); Kanda et al., Biotechnol Bioeng 94(4): 680-8 (2006)). Others have shown that binding low-fucose Herceptin to FcγRIIIa was improved by about 50-fold over normal-fucose Herceptin, and as a result, Herceptin-mediated-ADCC was substantially improved (Shields et al., J Biol Chem 277(30): 26733-40 (2002)). This indicates that defucosylation of anti-cancer antibodies increases their binding affinity to FcγR and enhances ADCC. Further, it suggests that modification of fucose content represents a way to improve anti-cancer immune response of antibodies so as to augment their therapeutic efficacy and expand the treatment to cancer patients that are unresponsive to fucosylated antibodies.
However, there remains a need for alternative methods for modifying antibodies of high therapeutic potential.